The rate card says $0.60 per 1,000. The invoice says something else. Three multipliers move the number without touching the headline price: which queue you pick, how deep you fetch, and whether you retry.
Get those right and the per-1,000 math is predictable. Miss one and a test budgeted at $0.60 bills many times higher. One buyer described a bill that came back roughly fifty times the budget once depth-based pricing kicked in, and the shape of that surprise is what this page takes apart.
This is the calculator. Whether to pay as you go or subscribe at all is a separate decision. Here the workload stays fixed and the request varies.
Every SERP API Quotes per 1,000, but Not per the Same SERP
One billing unit is one results page holding ten results. That is the first line to verify on any pricing page, because depth and queue both multiply that unit.
DataForSEO's published pay-as-you-go SERP pricing splits into three tiers for the same page:
- Standard queue: $0.60 per 1,000 SERPs. Queued: you post the task, poll for it, and get the result when the queue drains.
- Priority queue: $1.20 per 1,000. A faster queue than Standard.
- Live mode: $2.00 per 1,000. Real time, result in the same call.
Per request that is $0.0006, $0.0012 and $0.002. Priority is twice Standard. Live is 3.3 times Standard on the published list, and users who switched a workload from Standard to Live describe closer to a fivefold jump in that month's bill once retries and extra endpoints came along for the ride. The gap between list price and metered bill is where queue choice lives.
If speed does not matter (a bulk backfill, a weekly audit, training data) Standard is the sane default. If you need the page now, for an agent loop or a user waiting on a screen, you pay Live. Do not compare one provider's $0.60 Standard with another's $2.00 Live and call it a price difference. It is a latency tier difference.
Depth Is the Multiplier Buyers Miss
Depth is how many results you pull per keyword. Depth 10 is one page. Depth 100 is ten pages. Pricing pages list the first page, and depth-based models bill each additional page as its own SERP.
A September 2025 pricing note describes the DataForSEO shape plainly: $0.60 per 1,000 for the first page, with additional pages at $0.45 per 1,000. Treat the additional-page line as the shape to budget against and confirm it for the exact endpoint at checkout, because it is not consistently published for every mode.
One keyword at depth 10 is one SERP. One keyword at depth 100 is ten. You are not buying a thousand keywords. You are buying pages.
Worked Calculator: The Same 1,000 Keywords, Four Ways
Use this before you paste any per-1,000 number into a spreadsheet. The first-page tiers are the published $0.60, $1.20 and $2.00. The depth-100 rows use $0.45 per additional page for Standard; Live additional-page pricing is not consistently published, so the Live depth-100 row is illustrative and will be higher if your provider bills extra pages at the Live rate.
| Workload | Queue | Pages per keyword | Cost per keyword | 1,000 keywords | 100,000 keywords |
|---|---|---|---|---|---|
| Depth 10 | Standard | 1 | $0.0006 | $0.60 | $60.00 |
| Depth 10 | Live | 1 | $0.002 | $2.00 | $200.00 |
| Depth 100 | Standard | 10 | $0.00465 | $4.65 | $465.00 |
| Depth 100 | Live (illustrative) | 10 | $0.00605 | $6.05 | $605.00 |
Depth-100 per keyword is the first page plus nine additional pages: Standard $0.0006 + 9 × $0.00045 = $0.00465; Live illustrative $0.002 + 9 × $0.00045 = $0.00605.
Read the column you actually care about.
- Depth 10: Live is 3.3 times Standard ($2.00 against $0.60). That is the queue tax.
- Depth 100: even on Standard, depth 100 is 7.75 times depth 10 ($4.65 against $0.60). That is the depth tax.
- Both together: Live at depth 100 against Standard at depth 10 is about ten times.
Scaled to a monthly habit:
- 10,000 keywords a month, depth 10: Standard $6, Live $20.
- 10,000 keywords a month, depth 100: Standard about $46.50, Live illustrative about $60.50.
- A million keywords checked once: Standard depth 10 $600, depth 100 about $4,650. The $600 figure that circulates for a million lookups assumes depth 10. Change the depth and the total moves without any price change.
That is how a budget built on $0.60 ends up feeling fifty times off. It is rarely fifty pages against one on every keyword. Mix a deeper fetch with a faster queue and a retry policy, and the multipliers stack fast.
Size the Workload Before You Trust the Rate Card
Start from the workload, not the rate card.
- Count keywords per month.
- Decide depth per keyword. Do you need the top ten or the top hundred? Be literal: ten pages is ten SERPs.
- Pick a queue per use case: bulk gets Standard, user-facing gets Live.
- Multiply: keywords × pages × queue price.
Take 5,000 keywords, where half need depth 10 on Standard for a backfill and half need depth 100 on Live for an agent loop:
- 2,500 × 1 × $0.0006 = $1.50
- 2,500 × 10 pages, at the illustrative $0.00605 per keyword for Live depth 100 = $15.13
- About $16.63 for the month, before retries and before any wrapper fee or deposit floor.
Swap all 5,000 to Standard at depth 10 and the same workload is $3.00. The $13 difference is queue and depth, not provider markup.
If you want to validate that arithmetic before parking a deposit with a metered vendor, a small capped plan is the cheapest test bucket. qequals sells one $9 plan with a fixed monthly SERP allowance and no deposit, which is enough to check a depth and queue assumption against real pages before you commit to volume.
Checklist Before You Buy: Queue, Depth, Retries
1. Which queue will you actually use? Confirm turnaround and price for Standard, Priority and Live on the exact endpoint you will call (organic results, AI Overview, and so on). If you budget on Standard and ship on Live, expect three to five times in practice. If you can tolerate polling, Standard saves immediately.
2. What depth do you need, and how is it billed? Ask whether depth 100 is billed as ten SERPs, and at what per-page rate after the first. Get the line in writing. Then model depth 10, 20 and 100 as separate lines; one per-1,000 number cannot be extrapolated across them.
3. Do retries and failed fetches bill? Some providers bill a retry as a new SERP. If you poll a queued job twice or refetch on a stale parse, the meter runs again. Ask whether cache hits burn quota, because on some stacks they do, and whether the poll that collects a queued result counts as its own request.
4. How do you count results against requests? Verify that ten results equal one request in your provider's docs, and whether AI Overview or image results are separate endpoints with their own per-1,000 pricing. One number rarely covers every SERP type.
5. Will you dedupe before you fetch? If you re-request the same keyword weekly without a cache, you pay again each week. Plan deduplication on your side; do not assume the wrapper does it.
The deposit minimums, top-ups and expiring credits that sit around these numbers change the bill again, and they deserve their own pass. This page is the request-level arithmetic only: pick the queue that fits your latency, set depth to what you will actually parse, and let the per-1,000 column keep you honest.
